[R] variance components in lme

[Dennis L. Malandro]

Hello,

I'd tried to estimate the variance components for the
following data set, where lab and run %in% lab are
random effects, and vol is the response.

   lab run       vol
37   2   1 0.6800000
38   2   1 0.6300000
39   2   1 0.6600000
40   2   1 0.7600000
41   2   1 0.6800000
42   2   1 0.6100000
67   2   2 0.2940494
68   2   2 0.2943254
69   2   2 0.3339296
70   2   2 0.2439775
71   2   2 0.2516614
72   2   2 0.2281578

> lme(vol ~ 1, random = ~ 1 | lab/run)
Linear mixed-effects model fit by REML
  Data: NULL 
  Log-restricted-likelihood: 14.23677
  Fixed: vol ~ 1 
(Intercept) 
  0.4721751 

Random effects:
 Formula: ~1 | lab
        (Intercept)
StdDev:  0.03570064

 Formula: ~1 | run %in% lab
        (Intercept)  Residual
StdDev:   0.2791252 0.0463765

Number of Observations: 12
Number of Groups: 
         lab run %in% lab 
           1            2 


Since lab has only one level, I would have expected
StdDev for Formula: ~ | lab (Intercept) to be zero or
NaN or something like that, instead of 0.03570064. 
What am I misinterpreting?  Or better yet what is
0.03570064 the standard deviation estimate of, if not
that between labs?

Thanks a lot,
Dennis

Windows 95
R 1.4.1 (and the version of nlme that came with it)

__________________________________________________



-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._